this new AI model seamlessly integrates your product into any image without changing it
they call it nano-banana, and in this thread I’ll show you some examples, info and how to try it out for free 👇 https://t.co/a0BJgA47Oc

Lately there are rumors this model is just a preview of Google’s upcoming Gemini 3.
nano-banana is the nickname used in benchmarks, where it already outperforms models like Flux Kontext and GPT 5.
more examples.
1. 👈 nano-banana | GPT-5👉
fun fact: assets created on chatGPT https://t.co/PrOcaBEMYG




2. 👈 nano-banana | Flux 1. Kontext👉
prompt: put the woman lying over the car
The car stays the same, even the plate, with the woman lying on top. Kontext handles the character well but ruins the car.
(this time, ref images were created using Flux Kontext) https://t.co/zFHH25z1xv




3. 👈 nano-banana | Flux 1. Kontext👉
prompt: the woman is wearing the black mask
Look at the reflections… there’s no competition here. https://t.co/NXK4bLcWBc




4. 👈 nano-banana | Gemini 2👉
put the text and logo integrated on the first image
just perfect https://t.co/jb7ygfw929




The only way I know to try it is on @lmarena_ai:
switch to battle mode → image.
You’ll often see nano-banana results, and they’re impressive!
Try it here:
https://t.co/he9zzrgI3b https://t.co/vnzBEocK65

@lmarena_ai Hope you liked it!
Follow me at @techhalla for more guides, updates, and generative AI prompts.
Leave a like, repost, and bookmark along with your comments on the main post, it really helps!
https://t.co/i2FpwoTEQU